Mersenne Digest Wednesday, November 3 1999 Volume 01 : Number 655
----------------------------------------------------------------------
Date: Mon, 01 Nov 1999 19:35:03 -0800
From: Stefan Struiker <[EMAIL PROTECTED]>
Subject: Mersenne: MersenneInfo: What Happened?
To The MBase:
Haven't received any forwards since Friday Oct. 29th...
MersennnialMadMan,
Stefan S.
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Mon, 1 Nov 1999 23:33:18 -0500 (EST)
From: Lucas Wiman <[EMAIL PROTECTED]>
Subject: Re: Mersenne: MersenneInfo: What Happened?
> To The MBase:
>
> Haven't received any forwards since Friday Oct. 29th...
I believe the problem lies at the source. I don't think that anybody
has been sending mail to [EMAIL PROTECTED]
- -Lucas
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 2 Nov 1999 11:18:08 +0000
From: "Steinar H . Gunderson" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: MersenneInfo: What Happened?
On Mon, Nov 01, 1999 at 07:35:03PM -0800, Stefan Struiker wrote:
>Haven't received any forwards since Friday Oct. 29th...
That's because the list is so silent.
We need more life. Allow me :-)
Poaching is evil. The millennium starts year 2001. GIMPS is cooler that
distributed.net and SETI@Home combined. And of course, we shouldn't start
discussions like this ;-)
/* Steinar */
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 02 Nov 1999 13:04:07 -0500
From: "Glenn Maitz" <[EMAIL PROTECTED]>
Subject: Mersenne: Meganet Corp.
Dear everyone,
I may have missed this, but was there any final judgment on the
Meganet Corp. claim that they had a "deterministic and polynomial-time"
prime test? There were some discussions on the list early this year when
they first made their claim. But I don't recall reading about any
resolution. Did this just fade away? Are they still holding to their
claim? And if they are, does it stand up to examination (if they are
revealing enough for examination)?
Glenn Maitz
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 02 Nov 1999 12:57:13 -0600
From: "William H. Geiger III" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Meganet Corp.
In <[EMAIL PROTECTED]>, on 11/02/99
at 01:04 PM, "Glenn Maitz" <[EMAIL PROTECTED]> said:
>Dear everyone,
> I may have missed this, but was there any final judgment on the Meganet
>Corp. claim that they had a "deterministic and polynomial-time" prime
>test? There were some discussions on the list early this year when they
>first made their claim. But I don't recall reading about any resolution.
>Did this just fade away? Are they still holding to their claim? And if
>they are, does it stand up to examination (if they are revealing enough
>for examination)?
These guys are snake-oil vendors. I don't know what type of prime test
they are claiming to have or not have but from my exposure to their crypto
claims I wouldn't trust anything from them without proof.
- --
- ---------------------------------------------------------------
William H. Geiger III http://www.openpgp.net
Geiger Consulting Cooking With Warp 4.0
Author of E-Secure - PGP Front End for MR/2 Ice
PGP & MR/2 the only way for secure e-mail.
OS/2 PGP 5.0 at: http://www.openpgp.net/pgp.html
Talk About PGP on IRC EFNet Channel: #pgp Nick: whgiii
Hi Jeff!! :)
- ---------------------------------------------------------------
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 02 Nov 1999 14:50:53 -0500
From: Bill Daly <[EMAIL PROTECTED]>
Subject: Re: Mersenne: LL and Pollard-rho in one?
The behavior of the recursion x[n+1] = x[n]^2 - 2 can be precisely
analyzed. (In fact, it is because of this that the LL-test works at
all.) For a fixed prime p, the periodicity depends on the factorization
of either p-1 (if x[1]^2-4 is a quadratic residue mod p) or p+1 (if
x[1]^2-4 is a quadratic nonresidue mod p). In either case, let the
factorization be t*2^m where t is odd. Without going into excruciating
detail, the length of the non-periodic portion starting with x[1] will
be at most m, and the length of the loop will be a divisor of phi(t)/2
(the Euler totient function). (The behavior of the recursion x[n+1] =
x[n]^2 is similar, except that only the factorization of p-1 is
involved.) For example, when p = 41, the loops are {-1}, {2}, {6,-7},
{14,-11,-4} and {8,-20,-12,19,-10,16}. The value x[1] = 13 has the
non-periodic portion 13, 3, 7 of length 3 before entering the {6,-7}
loop, corresponding to the factorization 41-1 = 5*2^3. The lengths of
the loops {2} and {6,-7} are factors of phi(5)/2 = 2. Similarly, the
lengths of the loops {-1}, {14,-11,-4} and {8,...,16} are factors of
phi(21)/2 = 6.
>From the point of view of factorization, these periods are much too long
to be practical. If we use x[n+1] = x[n]^2 + k, where k <> 0 and k <>
- -2, then empirically the length of the loops mod p will be much smaller
than those for k = 0 or k = -2, while the length of the non-periodic
portion can easily be greater than m. For example, the recursion x[n+1]
= x[n]^2 - 6 mod 41 has three loops, {-2}, {3}, and {14,26}, and the
initial value x[1] = 0 has a non-periodic part of length 9. This is why
Pollard-rho works best for values of k other than 0 and -2.
An efficient Pollard-rho-like algorithm for factoring a Mersenne number
Mp is based on the iteration x[n+1] = x[n]^p + k. The idea is that
x[n+1] - x[n] = x[n]^p - x[n-1]^p = (x[n] - x[n-1]) * f[n], where f[n] =
(x[n]^p - x[n-1]^p) / (x[n] - x[n-1]) is known to have prime factors
which are all congruent to 1 mod p. Since Mp must also have prime
factors which are congruent to 1 mod p, this increases the chance that
gcd(x[n+1] - x[n], Mp) will produce a factor of Mp. Note that x[n+1] -
x[n] = f[n]*f[n-1]*f[n-2]*...*f[j]*(x[j] - x[j-1]), thus if any of the
terms in f[n]*f[n-1]*...*f[j] has a factor in common with Mp, this will
show up in the gcd. Ordinary Pollard-rho will also (given enough
iterations) find a factor of Mp, but the probability of this happening
within the first p-2 steps is minuscule. On the other hand, it would
cost little, after a failed LL-test, to calculate the gcd of the
difference of the last two results with Mp, just in case this happens to
reveal a factor. For that matter, one could just take the difference
between the final result and the last value stored in the p.... file.
Maybe if prime95 saved the final result in an r.... file, this test
could be performed later.
One interesting sidelight to this. The usual test for primality of a
Fermat number is to start with x[1] = 3, then calculate x[n+1] = x[n]^2
mod Fk up to x[2^k] mod Fk, which will be equal to -1 if and only if Fk
is in fact prime. Alternatively, one could start with y[1] = 3+1/3 mod
Fk (i.e., y[1] = (Fk+10)/3), then calculate y[n+1] = y[n]^2 - 2 mod Fk
up to y[2^k-1] mod Fk, which will be equal to 0 if and only if Fk is in
fact prime. This is because y[n] = x[n] + 1/x[n] mod Fk, thus y[2^k] =
(-1) + 1/(-1) = -2 mod Fk, thus y[2^k-1]^2 - 2 = y[2^k] = -2 mod Fk,
which implies y[2^k-1] = 0 mod Fk. If we calculate mod Fk(Fk+2) instead
of mod Fk, then the calculations can be carried out by prime95 with only
trivial modifications, and at the end we will have to reduce y[2^k-1]
mod Fk(Fk+2) to y[2^k-1] mod Fk to complete the test. Of course, there
are no longer any Fk within reach of current computers, so this is
merely of academic interest.
I suggested another (small) group of potential primes for testing in an
earlier message, but it never got through to the list (most likely
because my e-mail address has changed). Let F3(n) = (2^3^(n+1) - 1) /
(2^3^n - 1) = 2^(2*3^n) + 2^3^n + 1. F3(n) grows more quickly than the
Fermat numbers, so there aren't very many within reach of current
computers. There is a primality test for F3(n) which is similar to that
for Fermat numbers. 5 is a quadratic nonresidue of F3(n) for all n, thus
if F3(n) is prime, then 5^((F3(n)-1)/2) = -1 mod F3(n). On the other
hand, if 5^((F3(n)-1)/2) = -1 mod F3(n), and q is a divisor of F3(n),
then the congruence also holds mod q, which implies that q = 1 mod
2^3^n. However, since (2^3^n + 1)^2 > F3(n), it is not possible for
F3(n) to have two factors both of which are congruent to 1 mod 2^3^n,
thus if 5^((F3(n)-1)/2) = -1 mod F3(n), F3(n) must be prime. Thus, start
with x[1] = 5 and calculate x[j+1] = x[j]^2 mod 2^3^(n+1)-1 up to
x[3^n+1], then y[1] = 5*x[3^n+1] mod 2^3^(n+1)-1, then y[j+1] = y[j]^2
mod 2^3^(n+1)-1 up to y[3^n]. Reduce y[3^n] mod 2^3^(n+1)-1 to y[3^n]
mod F3(n). If the result is -1, then F3(n) is prime, otherwise F3(n) is
composite. (The idea of using 2^3^(n+1)-1 as a modulus is to be able to
take advantage of prime95's DWT code.) It's pretty unlikely that this
will find a large prime, but who knows?
Regards,
Bill
**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom
they are addressed.
This footnote also confirms that this email message has been swept by
MIMEsweeper for the presence of computer viruses.
**********************************************************************
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 2 Nov 1999 20:49:01 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Trial-factorers
On 27 Oct 99, at 17:23, Eric Hahn wrote:
>
> I'm looking for program(s) capable of trial-factoring
> prime exponent Mersenne numbers (using 2kp+1) meeting
> the following requirements:
>
> 1) Capable of trial-factoring any exponent > 1
> (at least to some considerably large number,
> say 1 trillion?)
>
> As I recall, Brian [Beesley] mentioned something once
> about having a program that could test an exponent
> of an arbitrary size... Brian??
Umm - I did write a "quick hack" which could handle exponents up to
1/6 * 2^32, but I never got round to putting a usable front end on
it. Actually I thought that the release of the Prime95 v19 beta made
it pretty redundant, and I was having a hard time with kidney stones.
>
> 2) Capable of testing a factor of any size.
> (even over the 2^76 limit of Prime95).
>
> I just know somebody is going to have to mention the
> time involved in testing factors of such a large size.
> Let me just say, I realize *exactly* how much time
> would be required...
>
> 3) Capable of trial-factoring a range of k's.
> (example: from k=1000 to k=2500)
Well, I'm prepared to have a go. Could we tighten up the spec a bit?
(a) There's also been some interest in something else that Prime95
doesn't do - trial factoring 2^p+1.
(b) I assume we're only interested in 2kp+1 factors. This means that
we will miss any factors which are not of this form. (Applies to
Mersenne numbers with composite exponents, and all 2^p+1 numbers -
though I believe that the "missed" exponents are easy to derive
analytically.)
(c) I presume we're looking for a program optimized for IA32
architecture. The mersfac* programs are available but are unlikely to
be optimally efficient on any particular hardware platform.
Given that, I suggest limits on exponent < 2^62 and on factor < 2^95
(these are convenient for the architecture).
It's probably sensible to go for an application which runs in a "DOS
box" rather than a proper windowed application. This makes it a bit
easier (for me) to write & also makes deriving a linux variant almost
trivial. (Does anyone know for sure whether or not there's a DOS box
in "Millenium"? I heard a nasty rumour...)
If we can agree on that, then I have a basis to begin coding. Will
certainly take a month or two to produce even a pre-pre-release as I
am very pushed for time at the moment.
Regards
Brian Beesley
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Tue, 2 Nov 1999 20:49:01 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Questions about prime.ini syntax, and hardware advice.
On 27 Oct 99, at 21:13, Albert Garrido wrote:
> I'm currently trying to configure the Time command, as listed in the
> docs, to
> get the prime95 client to function as follows.
>
> User ID=XYZABC
> Time=1-5/18:00-0:00,1-5/0:00-08:00,6-7/0:00-24:00
> (reset of Prime.ini)
>
> Anyhow, it doesn't seem to work, and I tried a few other syntax forms,
> but
> what would usually happen is that either the app runs at 6:00PM and then
> goes
> to sleep at 6:01. I inserted the line Priority=1 after the Time line,
> and it
> doesn't do much. I couldn't find an explanation of the time variables
> on the
> FAQ, I'm hoping someone's been down this road before.
This is slightly tricky & not very well documented - though there is
explanatory text in the "undoc.txt" file which comes with the
official v19 distribution (and also with late betas of v19).
The trick is:
(a) the "Time=" line(s) should appear at the end of the file;
(b) any "Priority=" or "DiskWriteTime=" directives must appear
_after_ the "Time=" line to which they apply.
This allows you to use different priority & disk write time at
different times of day, even if the program is running 24 x 7.
The downside is that you _cannot_ have "Priority=" or
"DiskWriteTime=" before the first "Time=" line; if you do, then the
time directives will not work as expected.
Also, the end point in your first time interval should be 24:00 not
0:00. The end time must always be greater than the start time.
I would suggest you resequence your Prime.INI file as follows:
[ all other lines]
Time=1-5/00:00-08:00
Time=1-5/18:00-24:00
Time=6-7/00:00-24:00
If you have any Priority= or DiskWriteTime= directives, delete them
from the "top section" & make copies after each of the Time=
directives.
This should run the program at all times except 08:00-18:00
(according to your system clock) Monday to Friday.
I'm doing something a bit similar, and it does work; the end of my
Prime.INI file looks like this:
Time=1-7/00:00-08:05
Priority=1
DiskWriteTime=30
Time=1-7/20:00-24:00
Priority=1
DiskWriteTime=30
[Runs the program from 20:00 to 08:05 7 days a week]
I know that Priority=1 & DiskWriteTime=30 are defaults & could
therefore be omitted. I was messing about!
> Anyhow, it's an Athlon that's running at 900Mhz, besides the novelty
> value of
> that. How fast would something like this get through an average LL
> test?
{Green with envy}
I don't know whether Kryotech does anything with the bus/memory
speed, if it's "just" a CPU speedup (i.e. a larger multiplier) then
the effect will be significantly less than 1.5 times as fast as a
"room temperature" Athlon @ 600 MHz. But then the Athlon should be
significantly faster than a PIII at the same bus & CPU speed - even
without re-tuning to take advantage of the Athlon architecture.
My guess is, if the Krytotech module is driving the Athlon at 9x100
MHz, the performance is probably going to be almost exactly twice
that of a PIII-450 - which is about 10% better than the "standard"
benchmark PII-400 - so multiply the benchmark figures by 0.45. Should
do better than that if the bus speed is higher than 100 MHz.
Quite apart from anything else, it would be interesting to see how
well Prime95 _does_ run on an Athlon. Perhaps you would care to run a
QA test suite (takes about 12 hours on a PIII-450). You might also be
interested in parallel running some tests with the QA team; the extra
interest here is that this is probably as good a way as any of making
a name for yourself by finding a flaw in the Athlon FPU !!!
Regards
Brian Beesley
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 03 Nov 1999 09:50:37 +1300 (NZDT)
From: Bill Rea <[EMAIL PROTECTED]>
Subject: Mersenne: Mlucas on SPARC
Gimpsters,
If there are people with SPARC systems with large external caches
thinking about using Ernst Mayer's Mlucas for LL testing then
make the switch. I will be switching from MacLucasUNIX on an E450
with 400Mhz CPU with 4Mb caches to Mlucas as soon as the current
exponents finish. I ran a few tests to check on the speed differences
with the current exponents and to my surprize on this system Mlucas
is significantly faster. At 256K FFT I was seeing 0.11 secs/iter with
Mlucas against 0.15 secs/iter for MLU, at 512K FFT the figures were
0.25 secs/iter and 0.29 secs/iter respectively. Mlucas uses far less
memory that MLU and with a 4Mb cache you will really benefit from the
larger cache size. I would add that this is using the 32-bit code
as well. Mlucas runs significantly faster if you can compile and run it
on a 64-bit Solaris 7 system. MLU is unusual in that it runs significantly
slower in 64-bit mode. On small cache systems, MLU is faster for
current exponents.
I would like to upgrade the E450 to Solaris 7 and see what Mayer's
code can do in 64-bit mode, but the owners won't let me :-(
Bill Rea, Information Technology Services, University of Canterbury \_
E-Mail b dot rea at its dot canterbury dot ac dot nz </ New
Phone 64-3-364-2331, Fax 64-3-364-2332 /) Zealand
Unix Systems Administrator (/'
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 00:39:15 -0500 (EST)
From: Lucas Wiman <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Trial-factorers
> Well, I'm prepared to have a go. Could we tighten up the spec a bit?
>
> (a) There's also been some interest in something else that Prime95
> doesn't do - trial factoring 2^p+1.
(Note that this form also form also has factors of the form k*p+1.)
> (b) I assume we're only interested in 2kp+1 factors. This means that
> we will miss any factors which are not of this form. (Applies to
> Mersenne numbers with composite exponents, and all 2^p+1 numbers -
> though I believe that the "missed" exponents are easy to derive
> analytically.)
Yes, those mersenne numbers with composite exponents have factors of
the form 2^d-1 where d|p, but the remaining unfactored portion must have
factors of the form k*p+1. (I believe that is Legendre's theorem)
(or rather a consequence of it)
> It's probably sensible to go for an application which runs in a "DOS
> box" rather than a proper windowed application. This makes it a bit
> easier (for me) to write & also makes deriving a linux variant almost
> trivial. (Does anyone know for sure whether or not there's a DOS box
> in "Millenium"? I heard a nasty rumour...)
Egad! If true then it would be added to the long list of complaints I have
with microsoft.
> If we can agree on that, then I have a basis to begin coding. Will
> certainly take a month or two to produce even a pre-pre-release as I
> am very pushed for time at the moment.
No need to rush, I wouldn't call the need for such a program urgent.
- -Lucas
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 07:22:06 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Mlucas on SPARC
On 3 Nov 99, at 9:50, Bill Rea wrote:
> If there are people with SPARC systems with large external caches
> thinking about using Ernst Mayer's Mlucas for LL testing then
> make the switch.
I actually switched my Ultra 10 to Mlucas (using the binary you
supplied for Solaris 2.6) a couple of weeks back. I don't know what
the cache size on this system is, but I doubt it's as big as 4MB.
The point is that I found (a) MLU at FFT size N is about the same
speed as Mlucas at FFT size 7N/8, but (b) this is offset to a large
extent by the fact that Mlucas runs larger exponents than MLU with
the same (power of 2) FFT size.
e.g. on exponents around 3.7 million the iteration time for Mlucas is
0.185 using 192K FFT size. MLU would need a 256K FFT size and would
run at 0.228 sec/iter. (Ultra IIi @ 300 MHz)
Of course, there will be _some_ exponents where MLU is a bit faster
on this particular system - so I'll keep MLU around.
Bear in mind that Mlucas is still a bit "experimental". One problem
has come to light - if you're using the PrimeNet Manual Testing page
to submit results, you _must_ remove the space preceding the M at the
beginning of the result line, else PrimeNet won't accept the result.
Actually it says it has got the result but doesn't remove the
assignment from the current assignments report or add it to the
completed assignments report unless the leading space is removed.
(This seems to be a program bug, I see the extra space in the source
code so I'd assume the problem afflicts Mlucas 2.7y on all platforms.
Certainly the binary executable I built for Alpha systems running
linux is also afflicted.)
> MLU is unusual in that it runs significantly slower in 64-bit mode.
Very odd. Any idea why this might be? Compiler producing broken
code???
Regards
Brian Beesley
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 03 Nov 1999 13:53:59 +0100
From: Alex Kruppa <[EMAIL PROTECTED]>
Subject: Mersenne: Factoring 2^n+1 and 2^n-1
Hi,
Ernst Mayer once mentioned to me that Prime95 needs twice the FFT size for 2^n+1
numbers (compared to 2^n+1 numbers). Does that mean that George is using the identity
2^(2n)-1 = (2^n+1)(2^n-1) ? I was wondering why ECM on 2^n+1 numbers took much longer
than on 2^n-1 of the same size..
That would mean that I can do ECM on, for example, P773 and M773 at the same time by
doing ECM on M1546, and it will take just as long as ECM on only P773, right? When I
find a factor, I'll just have see which number it divides.
Ciao,
Alex.
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 03 Nov 1999 13:57:44 +0100
From: Alex Kruppa <[EMAIL PROTECTED]>
Subject: Mersenne: Re: Factoring 2^n+1 and 2^n-1
Alex Kruppa wrote:
>
> Hi,
>
> 2^n+1 numbers (compared to 2^n+1 numbers).
^^^^^
Ack! Typos, I _hate_ them! Must be 2^n-1 of course.
Ciao,
Alex.
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 16:42:13 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Factoring 2^n+1 and 2^n-1
On 3 Nov 99, at 13:53, Alex Kruppa wrote:
> Ernst Mayer once mentioned to me that Prime95 needs twice the FFT size for 2^n+1
>numbers (compared to 2^n+1 numbers). Does that mean that George is using the identity
> 2^(2n)-1 = (2^n+1)(2^n-1) ? I was wondering why ECM on 2^n+1 numbers took much
>longer than on 2^n-1 of the same size..
It used to be the case that 2^n+1 was only using power-of-two FFT run
lengths. v19 has addressed this problem; I think you'll find that ECM
on 2^n+/-1 takes about the same time now (for the same n).
> That would mean that I can do ECM on, for example, P773 and M773 at the same time by
>doing ECM on M1546, and it will take just as long as ECM on only P773, right? When I
> find a factor, I'll just have see which number it divides.
Interesting idea ... but, the algorithm being O(n log n), it _should_
take a bit longer to run a single transform with run length 2N than
it does to run two transforms each with run length N.
This is particularly relevant since there is a good chance (with
small exponents) that the FFT will run from the processor cache; if
doubling the FFT run length causes overspill into main memory, the
performance could suffer dramatically.
Regards
Brian Beesley
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 15:43:24 +0100
From: "Steinar H. Gunderson" <[EMAIL PROTECTED]>
Subject: Mersenne: Re: Trial-factorers
On Tue, Nov 02, 1999 at 08:49:01PM -0000, Brian J. Beesley wrote:
>(Does anyone know for sure whether or not there's a DOS box
>in "Millenium"? I heard a nasty rumour...)
Don't know for sure. The problem to me, is finding out whether there
will _be_ a `Millennium' (based on 95/98 `technology'), and in that
case, if there will be a `Windows 2000' (based on NT) as _well_...
I've heard that Windows 2000 will have both COMMAND.COM and a CMD.EXE,
though.
Anybody care to port bash to Windows? (Oh well, it has probably been
done already...)
/* Steinar */
- --
Homepage: http://members.xoom.com/sneeze/
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 11:38:54 -0700
From: "Blosser, Jeremy" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Re: Trial-factorers
Yeh, W2K has both cmd.exe and command.com. Odd thing is that command.com
displays my typing REALLY REALLY slowly.
As far as 'Millenium'. I can't imagine there not being a DOS box. But then
again, I've also heard that it might not ever come out.
Lastly, bash has been ported to windoze already by the Cygnus folks... even
better tho (for NT at least), is 4NT (JPSoft).
- -----Original Message-----
From: Steinar H. Gunderson [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, November 03, 1999 8:43 AM
To: [EMAIL PROTECTED]
Subject: Mersenne: Re: Trial-factorers
On Tue, Nov 02, 1999 at 08:49:01PM -0000, Brian J. Beesley wrote:
>(Does anyone know for sure whether or not there's a DOS box
>in "Millenium"? I heard a nasty rumour...)
Don't know for sure. The problem to me, is finding out whether there
will _be_ a `Millennium' (based on 95/98 `technology'), and in that
case, if there will be a `Windows 2000' (based on NT) as _well_...
I've heard that Windows 2000 will have both COMMAND.COM and a CMD.EXE,
though.
Anybody care to port bash to Windows? (Oh well, it has probably been
done already...)
/* Steinar */
- --
Homepage: http://members.xoom.com/sneeze/
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 10:50:47 -0800
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Re: Trial-factorers
First, the important disclaimer, necessary because of my posting address: I
do not in any way speak for Microsoft in what I write below, but only in a
personal capacity.
In the W2k betas that have been issued so far, there is a "CMD" command
which does pretty much what you'd expect.
bash has been ported to NT long ago. It was done by Cygnus and is part of
the cygwin package. It works for me under W2k beta, but YMMV.
Paul
> -----Original Message-----
> From: Steinar H. Gunderson [mailto:[EMAIL PROTECTED]]
> Sent: 03 November 1999 14:43
> To: [EMAIL PROTECTED]
> Subject: Mersenne: Re: Trial-factorers
>
>
> On Tue, Nov 02, 1999 at 08:49:01PM -0000, Brian J. Beesley wrote:
> >(Does anyone know for sure whether or not there's a DOS box
> >in "Millenium"? I heard a nasty rumour...)
>
> Don't know for sure. The problem to me, is finding out whether there
> will _be_ a `Millennium' (based on 95/98 `technology'), and in that
> case, if there will be a `Windows 2000' (based on NT) as _well_...
> I've heard that Windows 2000 will have both COMMAND.COM and a CMD.EXE,
> though.
>
> Anybody care to port bash to Windows? (Oh well, it has probably been
> done already...)
>
> /* Steinar */
> --
> Homepage: http://members.xoom.com/sneeze/
> _________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
>
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 14:12:51 -0500
From: Pierre Abbat <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Trial-factorers
>Anybody care to port bash to Windows? (Oh well, it has probably been
>done already...)
It has been done, and you need only operate swans.
phma
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 11:36:24 -0800
From: "John R Pierce" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Trial-factorers
> As far as 'Millenium'. I can't imagine there not being a DOS box. But then
> again, I've also heard that it might not ever come out.
fyi, millenium HAS dos in-a-window, it just won't have a realmode boot mode
(standalone dos). they are stripping the realmode stuff down to be JUST a
boot loader rather than a full blown msdos kernel...
- -jrp
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 03 Nov 1999 15:08:15 -0500
From: Albert Garrido <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Trial-factorers
"Blosser, Jeremy" wrote:
>
> Yeh, W2K has both cmd.exe and command.com. Odd thing is that command.com
> displays my typing REALLY REALLY slowly.
Blame that on command.com running as a protected mode application
under Windows on Windows. It's only around to keep legacy
compatibility. For amusement? Watch what happens to the processor
utilization when you type.
I don't know. A microsoft snafu, or just the most expedient method of
making it work, under the circumstances.
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 17:06:46 -0700
From: "Aaron Blosser" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Re: Trial-factorers
command.com and cmd.exe have existed in NT since version 3.51, probably in
NTAS (ver 3.1) also.
Command.com runs, well, dog slow. It's 16 bit methinks, uses LOTS of cpu
time and is generally pretty lousy.
Furthermore, it doesn't support the cmd extensions that cmd.exe does.
CMD.EXE is 32 bit, is easier on the cpu and supports extensions. But, just
run any 16 bit app from within cmd.exe and you'll still invoke the ntvdm (or
is that the wowexec?). F'rinstance, run edit in a cmd shell and you'll
invoke wowexec or ntvdm (whichever...I can't remember). Once it's invoked,
it'll stay running until you reboot, but the first time it loads, it can
take a VERY long time depending on your network configuration of all things.
Running command.com will do this, whereas at least if you run cmd.exe and do
some simple stuff like: net start "prime service", you're not invoking the
16 bit command processor.
Whew.
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Blosser,
> Jeremy
> Sent: Wednesday, November 03, 1999 11:39 AM
> To: [EMAIL PROTECTED]
> Subject: RE: Mersenne: Re: Trial-factorers
>
>
> Yeh, W2K has both cmd.exe and command.com. Odd thing is that command.com
> displays my typing REALLY REALLY slowly.
>
> As far as 'Millenium'. I can't imagine there not being a DOS box. But then
> again, I've also heard that it might not ever come out.
>
> Lastly, bash has been ported to windoze already by the Cygnus
> folks... even
> better tho (for NT at least), is 4NT (JPSoft).
>
> -----Original Message-----
> From: Steinar H. Gunderson [mailto:[EMAIL PROTECTED]]
> Sent: Wednesday, November 03, 1999 8:43 AM
> To: [EMAIL PROTECTED]
> Subject: Mersenne: Re: Trial-factorers
>
>
> On Tue, Nov 02, 1999 at 08:49:01PM -0000, Brian J. Beesley wrote:
> >(Does anyone know for sure whether or not there's a DOS box
> >in "Millenium"? I heard a nasty rumour...)
>
> Don't know for sure. The problem to me, is finding out whether there
> will _be_ a `Millennium' (based on 95/98 `technology'), and in that
> case, if there will be a `Windows 2000' (based on NT) as _well_...
> I've heard that Windows 2000 will have both COMMAND.COM and a CMD.EXE,
> though.
>
> Anybody care to port bash to Windows? (Oh well, it has probably been
> done already...)
>
> /* Steinar */
> --
> Homepage: http://members.xoom.com/sneeze/
> _________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
> _________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
>
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
Date: Wed, 3 Nov 1999 22:19:02 +0200
From: =?iso-8859-1?Q?Ernest_L=F6tter?= <[EMAIL PROTECTED]>
Subject: Mersenne: Windows 2000/Millennium ...
This is a multi-part message in MIME format.
- ------=_NextPart_000_009A_01BF2649.6F168C40
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
I hope this is not off-topic, but I saw that this is under discussion =
...
I am running a beta of Windows 2000, which is based on NT 'technology', =
that being the more protective kernel and NTFS file system, etc. It does =
have a Command Prompt, as Windows NT 4 had, which is basically the same =
as the well-known MS-DOS Prompt. Windows 2000 (and NT4 for that matter) =
doesn't support booting to pure-DOS mode, however.
As far as I know, 'Windows Millennium' is in beta testing, which will be =
a continuation of the 'old tech' Windows 95, Windows 98 line, which will =
surely have the same MS-DOS Prompt as they had, and I cannot see that =
Millennium will differ from 95/98 in any major way which will influence =
DOS apps which ran on 95/98. So, as far as I know, there _will_ be a =
'Windows Millennium', which should arrive some time after Windows2000.
Relating to the question about COMMAND.COM and CMD.EXE, I looked at my =
installation, and found the following (Win2000 installed in C:\WINNT) :
CMD.EXE and COMMAND.COM in C:\WINNT\system32 : which does the following =
when I run it :
COMMAND.COM gives :
Microsoft(R) Windows NT DOS
(C)Copyright Microsoft Corp 1990-1996.
CMD.EXE gives :
Microsoft Windows 2000 [Version 5.00.2031]
(C) Copyright 1985-1999 Microsoft Corp.
Hopefully this is not too much off-topic rambling, but I hope it helps =
:)
- ------=_NextPart_000_009A_01BF2649.6F168C40
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.2516.1900" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>I hope this is not off-topic, but I saw =
that this=20
is under discussion ...</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>I am running a beta of Windows 2000, =
which is based=20
on NT 'technology', that being the more protective kernel and NTFS file =
system,=20
etc. It does have a Command Prompt, as Windows NT 4 had, which is =
basically the=20
same as the well-known MS-DOS Prompt. Windows 2000 (and NT4 for that =
matter)=20
doesn't support booting to pure-DOS mode, however.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>As far as I know, 'Windows Millennium' =
is in beta=20
testing, which will be a continuation of the 'old tech' Windows 95, =
Windows 98=20
line, which will surely have the same MS-DOS Prompt as they had, and I =
cannot=20
see that Millennium will differ from 95/98 in any major way which will =
influence=20
DOS apps which ran on 95/98. So, as far as I know, there _will_ be a =
'Windows=20
Millennium', which should arrive some time after =
Windows2000.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>Relating to the question about =
COMMAND.COM and=20
CMD.EXE, I looked at my installation, and found the following (Win2000 =
installed=20
in C:\WINNT) :</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>CMD.EXE and COMMAND.COM in =
C:\WINNT\system32 :=20
which does the following when I run it :</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>COMMAND.COM gives :<BR>Microsoft(R) =
Windows NT=20
DOS<BR>(C)Copyright Microsoft Corp 1990-1996.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>CMD.EXE gives :</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Microsoft Windows 2000 [Version =
5.00.2031]<BR>(C)=20
Copyright 1985-1999 Microsoft Corp.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>Hopefully this is not too much =
off-topic rambling,=20
but I hope it helps :)</FONT></DIV>
<DIV> </DIV>
<DIV> </DIV></BODY></HTML>
- ------=_NextPart_000_009A_01BF2649.6F168C40--
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
------------------------------
End of Mersenne Digest V1 #655
******************************
